Download - Topological Properties of Benzenoid Systems. IX *. On the Sextet …zfn.mpdl.mpg.de/data/Reihe_A/37/ZNA-1982-37a-0069.pdf · 2018. 2. 9. · IX. On the Sextet Polynomial 71 system

Transcript
  • This work has been digitalized and published in 2013 by Verlag Zeitschrift für Naturforschung in cooperation with the Max Planck Society for the Advancement of Science under a Creative Commons Attribution4.0 International License.

    Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschungin Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung derWissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht:Creative Commons Namensnennung 4.0 Lizenz.

    Topological Properties of Benzenoid Systems. IX *. On the Sextet Polynomial Ivan Gutman Faculty of Science, University of Kragujevac, Yugoslavia

    Z. Naturforsch. 37 a, 69—73 (1982); received September 26, 1981

    A number of mathematical relations for the sextet polynomial are derived. A graph has been introduced (the so called C-graph), representing those properties of a benzenoid system which are essential in the sextet theory of Clar. The main structural properties of the C-graph are deter-mined. The obtained results contribute towards a better understanding of the algebraic and combinatorial background of Clar's theory of the aromatic sextet.

    Introduction

    Few years ago Hosoya and Yamaguchi [1] intro-duced the sextet polynomial of a benzoid system B as

  • 70 I. Gutman • Topological Properties of Benzenoid Systems. IX. On the Sextet Polynomial

    C C O ^ Bj B 3

    coooo B,

    o—o—o—o-o A „ C

  • I. Gutman • Topological Properties of Benzenoid Systems. IX. On the Sextet Polynomial 71

    system contains triangles unless the C-graph and the inner dual coincide.

    12. The C-graph may also contain larger cycles, but then inside every fc-membered cycle (k ^4) there are at least k — 2 triangles. Every vertex belonging to a &-membered cycle (k^Z 4) belongs also to a triangle.

    13. The C-graph of a branched cata-condensed system contains a vertex of degree three which does not lie on a triangle (and therefore does not lie on any cycle).

    This latter statement follows from the observa-tion that the three hexagons which are adjacent to a branched hexagon of a branched cata-con-densed system are always mutually resonant.

    The properties 11 — 13 are important for the application of Beineke's theorem by which one can decide whether a C-graph is a line graph (see later).

    The C-graph represents those relations between the hexagons of a benzenoid system which are fundamental from the point of view of Clar's reso-nant sextet theory. Therefore C(B) could be per-haps named the Clar graph of the benzenoid sys-tem B.

    Some Graphic Polynomials and Their Relation to the Sextet Polynomial

    In order to formulate the dependence of the sextet polynomial on the C-graph we need some further definitions. Let G be a graph with n vertices and m edges. Two vertices of G are said to be inde-pendent if they are not adjacent. Two edges of G are said to be independent if they are not incident.

    Let the number of selections of k mutually inde-pendent vertices in G be denoted by o(G, k), where-as the number of selections of k mutually indepen-dent edges be denoted by p(G, k), k= 1,2, ... . In addition we shall adopt the convention

    o(G, 0) == p(G, 0) = 1 for all graphs G.

    According to the above definitions, o(G, 1) = number of vertices of G, and p(G, 1) = number of edges of G.

    Let us now define the graphic polynomials co (G) and n (G) as follows.

  • 72 I. Gutman • Topological Properties of Benzenoid Systems. IX. On the Sextet Polynomial

    o(C(B), k) = n. Taking this into account we arrive to the formula (1).

    Although Theorem 1 is a more or less direct consequence of the definition of C(B) and a>(G), we find this result important. Namely, Eq. (1) pro-vides a connection between the basic concepts of Clar's theory and the purely graph theoretical co-polynomial. In other words, Theorem 1 provides a precisely defined graph theoretical basis for Clar's resonant sextet theory.

    If G is the line graph of a graph H, we shall write either G = L(H) or H = £-i(G). If G is not a line graph, then i _ 1 ( G ) does not exist. For example, C(BÖ), C(BÖ) and C(B?) are not line graphs (i.e. L _ 1 (C(BF)) , i = 5, 6, 7 do not exist), whereas C(BT), i = 1, 2, 3, 4 and 8 are line graphs. Using standard graph theoretical methods [21] it is not difficult to construct L~1(C). Some examples are given in Figure 2.

    Theorem 2. If C = C (B) is a line graph, then a (B, x) = n (L_1 (C), x) — (n — WR) X

    and x»o{B, z-2) = a(L-i(C), x ) - ( n - nR)xN~2,

    where N is the number of vertices of L~1(C). In particular, if all the hexagons of B are resonant, then cr(B) and 7r(Z/-1(C)) coincide.

    Proof. Combine Theorem 1 with Lemma 1. Theorem 3. If all the hexagons of the benzenoid

    system B are resonant and its C-graph is a line graph, then all the zeros of the sextet polynomial are real and negative.

    Proof. Combine Theorem 2 with Lemma 2. The above result is of great importance for the

    calculation of Aihara's resonance energy [4—6], namely it guarantees its reality. If C(B) is not a line graph, then the zeros of the sextet polynomial

    A — X LI1 (C(B,)) nVCtej» L-'(C(B3))

    •¥• 4 C1 (C(BJ) C1 (C(B8))

    Fig. 2. Graphs whose line graphs are given on Figure 1.

    may be complex, what has already been noticed in [4],

    Theorem 4. If L~1(C(B)) exists and is acyclic, then

    xN a(B, x~2) =

  • 73 I. Gutman • Topological Properties of Benzenoid Systems. IX.

    Proof of Theorem 5 is based on the application of Beineke's theorem. If B is a branched cata-condensed system, then according to property 13, C(B) contains a triplet of mutually non-adjacent vertices which are all adjacent to a fourth vertex. From Beineke's theorem we know that this is a forbidden subgraph for line graphs. Hence C(B) is not a line graph.

    If B is a non-branched cata-condensed system, then C(B) is composed of linearly arranged blocks, each block being a complete graph. Thus none of the nine forbidden subgraphs is contained in C(B) and hence C(B) is a line graph.

    The construction of L - 1(C(B)) in this latter case is now straightforward.

    Concluding this paper we would like to point at some problems for which in the present moment we are not able to obtain a general solution. First, under which conditions are the graphs C(B) and D (B) isomorphic ? Second, under which conditions is C(B) a line graph ? A third related question is the characterization of those benzenoid systems for which L'1 (C) is acyclic.

    In this work all these problems have been fully solved for cata-condensed, but only partially for peri-condensed benzenoid systems.

    Appendix

    Whereas the mathematical properties of jr (G) have been extensively investigated [14—20], it seems that the co-polynomial has not yet been considered in the literature. Nevertheless, problems concerning the sets of independent vertices of a graph belong to the classical topics of graph theory.

    [1] H. Hosoya and T. Yamaguchi, Tetrahedron Letters, 1975, 4659.

    [2] E. Clar, The Aromatic Sextet, Wiley, London 1972. [3] W . C. Herndon and M. L. Ellzey, J. Amer. Chem. Soc.

    96, 6631 (1974). [4] J. Aihara, Bull. Chem. Soc. Japan 50, 2010 (1977). [5] Part IV: I. Gutman, Bull. Chem. Soc. Japan 51, 2729

    (1978). [6] Part V : I. Gutman, Z. Naturforsch. 33a, 840 (1978). [7] N. Ohkami, A. Motoyama, T. Yamaguchi, H. Hosoya,

    and I. Gutman, Tetrahedron 37, 1113 (1981). [8] I. Gutman, Match (Mülheim) 11, 127 (1981). [9] Part II: I. Gutman, Theor. Chim. Acta 45, 309 (1977).

    [10] A. T. Balaban and F. Harary, Tetrahedron 24, 2505 (1968).

    [11] 0 . E. Polansky and D. H. Rouvray, Match (Mül-heim) 2, 63 (1976).

    [12] 0 . E. Polansky and D. H. Rouvray, Match (Mül-heim) 2, 197 (1976).

    On the Sextet Polynomial

    Furthermore, a chemical application of the num-bers o(G, k) has been recently reported [23]. In the following some elementary facts about the co-poly-nomial will be mentioned.

    1. Let On, K n and K^ & be the graph with n ver-tices and without edges, the complete graph with n vertices, and the complete bipartite graph with a-\-b vertices, respectively. Then

    co(Ow) = (1 + x)n> co(K„) = 1 + nx,

    co( Ka,») = ( l+aO* + ( l + a O » - l . 2. If the graph G is composed of two disconnected

    parts Ga and G&, then co(G) = co(Ga)co(Gft). 3. Let v be a vertex of G and let Av be the set

    containing the vertex v and all the vertices ad-jacent to v. Then we have the recurrence relation

    co(G) = co(G - v) + xo)(G - Av).

    This formula enables a straightforward evalua-tion of the co-polynomial. Having in mind Theo-rem 1, the same result provides a recursive tech-nique for the computation of the sextet polynomials. 4. (d/dz)co(G) = 2 c o ( G - A „ ) ,

    V

    with the summation going over all vertices of the graph G. 5. co(L(G)) = ?t(G).

    6. If G is a line graph, then all the zeros of co(G) are real and negative. If, however, G is not a line graph, then it can occur that the zeros of co (G) are complex.

    A more detailed study on the co-polynomial will be reported elsewhere [24].

    [13] Part I : I. Gutman, Croat. Chem. Acta 46, 209 (1974). [14] H. Hosoya, Bull. Chem. Soc. Japan 44, 2332 (1971). [15] H. Hosoya, Theor. Chim. Acta 25, 215 (1972). [16] I. Gutman and H. Hosoya, Theor. Chim. Acta 48, 279

    (1978). [17] I. Gutman, Match (Mülheim) 6, 75 (1979). [18] E. J. Farrell, J. Comb. Theory B 27, 75 (1979). [19] C. D. Godsil and I. Gutman, Z. Naturforsch. 34 a, 776

    (1979). [20] C. D. Godsil and I. Gutman, J. Graph. Theory 5, 137

    (1981). [21] F. Harary, Graph. Theory, Addison-Wesley, Reading

    1969, Chapter 8. [22] D. Cvetkoviö, M. Doob, and H. Sachs, Spectra of

    Graphs, Academic Press, New York 1980. [23] R. E. Merrifield and H. Simmons, Theor. Chim. Acta

    55, 55 (1980). [24] F. Haraiy and I. Gutman, in preparation.